|Annu. Rev. Astron. Astrophys. 2002. 40:
Copyright © 2002 by Annual Reviews. All rights reserved
2.1. Radio halos
Over 40 years ago Large (1959) discovered a radio source in the Coma cluster that was extended even when observed with a 45' beam. This source (Coma C) was studied by Willson (1970) who found that it had a steep spectral index and could not be made up of discrete sources, but instead was a smooth "radio halo" with no structure on scales less than 30'. Willson further surmised that the emission mechanism was likely to be synchrotron, and if in equipartition required a magnetic field strength of 2 µG. In Fig. 1 we show the best image yet obtained of the radio halo in the Coma cluster. Other radio halos were subsequently discovered, although the number known remained under a dozen until the mid-90s (Hanisch 1982).
Figure 1. WSRT radio image of the Coma cluster region at 90 cm, with angular resolution of 55'' × 125'' (HPBW, RA × DEC) from Feretti et al (1998). Labels refer to the halo source Coma C and the relic source 1253+275. The grey scale range displays total intensity emission from 2 to 30 mJy/beam while contour levels are drawn at 3, 5, 10, 30, and 50 mJy/beam. The bridge of radio emission connecting Coma C to 1253+275 is resolved and visible only as a region with an apparent higher positive noise. The Coma cluster is at a redshift of 0.023, such that 1' = 27 kpc.
Using the Northern VLA Sky Survey (NVSS; Condon et al. 1998) and X-ray selected samples as starting points (Giovannini & Feretti 2000, Giovannini, Tordi & Feretti 1999) have performed moderately deep VLA observations (integrations of a few hours) which have more than doubled the number of known radio halo sources. Several new radio halos have also been identified from the Westerbork Northern Sky Survey (Kempner & Sarazin 2001). These radio halos typically have sizes ~ 1 Mpc, steep spectral indices ( < -1), low fractional polarizations (< 5%), low surface brightnesses (~ 10-6 Jy arcsec-2 at 1.4 GHz), and centroids close to the cluster center defined by the X-ray emission.
A steep correlation between cluster X-ray and radio halo luminosity has been found, as well as a correlation between radio and X-ray surface brightnesses in clusters (Liang et al. 2000, Feretti et al. 2001, Govoni et al. 2001a). A complete (flux limited) sample of X-ray clusters shows only 5% to 9% of the sources are detected at the surface brightness limits of the NVSS of 2.3 mJy beam-1, where the beam has FWHM = 45'' (Giovannini & Feretti 2000, Feretti et al. 2001). But this sample contains mostly clusters with X-ray luminosities < 1045 erg s-1. If one selects for clusters with X-ray luminosities > 1045 erg s-1, the radio detection rate increases to 35% (Feretti et al. 2001, Owen, Morrison & Vogues 1999). Likewise, there may be a correlation between the existence of a cluster radio halo and the existence of substructure in X-ray images of the hot cluster atmosphere, indicative of merging clusters, and a corresponding anti-correlation between cluster radio halos and clusters with relaxed morphologies, e.g., cooling flows (Govoni et al. 2001a), although these correlations are just beginning to be quantified (Buote 2001).
Magnetic fields in cluster radio halos can be derived assuming a minimum energy configuration for the summed energy in relativistic particles and magnetic fields (Burbidge 1959), corresponding roughly to energy equipartition between fields and particles. The equations for deriving minimum energy fields from radio observations are given in Miley (1980). Estimates for minimum energy magnetic field strengths in cluster halos range from 0.1 to 1 µG (Feretti 1999). One of the best studied halos is that in Coma, for which Giovannini et al. (1993) report a minimum energy magnetic field of 0.4 µG. These calculations typically assume k = 1, = 1, low = 10 MHz, and high = 10 GHz, where k is the ratio of energy densities in relativistic protons to that in electrons, is the volume filling factor, low is the low frequency cut-off for the integral, and high is the high frequency cut-off. All of these parameters are poorly constrained, although the magnetic field strength only behaves as these parameters raised to the 2/7 power. For example, using a value of k ~ 50, as observed for Galactic cosmic rays (Meyer 1969), increases the fields by a factor of three.
Brunetti et al. (2001a) present a method for estimating magnetic fields in the Coma cluster radio halo independent of minimum energy assumptions. They base their analysis on considerations of the observed radio and X-ray spectra, the electron inverse Compton and synchrotron radiative lifetimes, and reasonable mechanisms for particle reacceleration. They conclude that the fields vary smoothly from 2 ± 1 µG in the cluster center, to 0.3 ± 0.1 µG at 1 Mpc radius.
2.2. Radio relics
A possibly related phenomena to radio halos is a class of sources found in the outskirts of clusters known as radio relics. Like the radio halos, these are very extended sources without an identifiable host galaxy (Fig. 1). Unlike radio halos, radio relics are often elongated or irregular in shape, are located at the cluster periphery (by definition), and are strongly polarized, up to 50% in the case of the relic 0917+75 (Harris et al. 1993). As the name implies, one of the first explanations put forth to explain these objects was that these are the remnants of a radio jet associated with an active galactic nucleus (AGN) that has since turned off and moved on. A problem with this model is that, once the energy source is removed, the radio source is expected to fade on a timescale << 108 years due to adiabatic expansion, inverse Compton, and synchrotron losses (see Section 4.1). This short timescale precludes significant motion of the host galaxy from the vicinity of the radio source.
A more compelling explanation is that the relics are the result of first order Fermi acceleration (Fermi I) of relativistic particles in shocks produced during cluster mergers (Ensslin et al. 1998), or are fossil radio sources revived by compression associated with cluster mergers (Ensslin & Gopal-Krishna 2001). Equipartition field strengths for relics range from 0.4 - 2.7 h502/7 µG (Ensslin et al. 1998). If the relics are produced by shocks or compression during a cluster merger, then Ensslin et al. (1998) calculate a pre-shock cluster magnetic field strength in the range 0.2-0.5 µG.